Introduction
Phased array ultrasonic testing (PAUT) is presented in detail in the American Welding Society (AWS) D1.5 Bridge Welding Code (BWC). The code book presents the usual sections including Scope, Definitions, Personnel Requirements, Equipment, etc. In this article, the focus is on phased array ultrasonic testing (PAUT) accept/reject criteria and how to apply them real data from a structural steel welds.
AWS Bridge Welding Code PAUT Acceptance Criteria
The BWC defines four different discontinuity classes: A, B, C, and D. Discontinuity Class A is the most severe, intermediate Classes B and C, and least severe Class D. They are categorized by the ultrasonic wave reflected amplitude of the weld discontinuity detected by phased array. The four different amplitude threshold levels are Automatic Reject Level (ARL), Standard Sensitivity Level (SSL), and Disregard Level (DRL). The SSL is set using the 0.060” diameter side drilled hole (SDH) on the IIW calibration block to 50% full-screen height (%FSH). The ARL is set 5 dB above the SSL and which is equivalent to 89 %FSH. Lastly, the DRL is set to 25 % FSH. The phased array discontinuity classes are defined in the AWS D1.5 Discontinuity Classification table shown below.
In addition to ultrasonic testing discontinuity classification, the BWC also defines PAUT minimum discontinuity lengths by types of by discontinuity classification. The allowable lengths are further classified by type of loading: compression and tension. For compression connections, the maximum lengths for Classes A, B, and C are non-allowed, 20 mm, and 50 mm, respectively. For tension joints, the maximum lengths for Classes A, B, and C are non-allowed, 12 mm, and 50 mm (middle) and 20 mm (top and bottom quarters), respectively. These criteria are outlined in the PAUT Acceptance Criteria outlined below.
In addition to the rules outlined in the PAUT Discontinuity Classification and Acceptance Criteria tables, there are some additional discontinuity separation criteria that must also be considered. Classes B and C reflections must be separated by at least 2 times the length of the longer indication. Additionally, Class B and C indications must not initiate at a distance less than 2 times the length from the weld end.
Example Phased Array Ultrasonic Testing Setup and Data
Figure 1 shows example data from a 1” thick steel plate with a single-vee joint and backing bar. The standard sensitivity level was set to 50 % FSH using the 0.060” side drilled hole on the IIW block. On the Olympus MX-2 platform, this step is accomplished during the sensitivity calibration step. During this step, compensation gain is added to each focal law that accommodates the attenuation characteristics of each phased array focal law or beam. In addition, a secondary sizing calibration step was performed to establish a TCG/DAC function that normalizes the amplitude response from the 0.060 side drilled hole over the desired testing range. Typically, a NAVSEA or comparable block, is used for this step.
During the TCG/DAC setup process using the Olympus MX-2 platform, additional curves may be established that serve as useful interpretation tools. The A-scan in the top left-hand side of phased array data shown Figure 1 has three different sensitivity levels superimposed onto PAUT A-scan data. The DRL is the black horizontal line at 25 %FSH, SSL the red line at 50 %FSH, and the ARL is the black line at 89 %FSH.
The PAUT encoded C-scan data shown below identified three roundish reflectors that were detected in the second leg. The first indication is observed at approximately 8.6” from the left edge. The second and third indications are located at 10.75” and 12” from the left edge, respectively. The blue C-scan data cursor is shown at 10.75”. Using the C-scan and S-scan data cursors, the maximum reflection amplitude is identified and 6 db sizing is used to measure the length of the middle indication. Using this technique this middle indication was determined to be approximately 0.75” long. The measured A-scan amplitude is 30 %FSH and this PAUT indication is therefore classified as a Class C discontinuity since it registers in between the DRL and SSL levels. Assuming this joint is in compression, this individual indication would pass since the length is less than 50 mm or 2”. Upon further investigation, however, it is observed that the indications to the left and to the right are within the 2 times the discontinuity length criteria outlined above. Using the PAUT C-scan, the middle indication at 10.75 inches was measured to be approximately 0.75”. The center of the PAUT indication to the right is located at approximately 12”. The two indications are clearly inside the 2 times the discontinuity length criteria (1.5 inches).
The PAUT indication on the left is also within the 2 times the discontinuity length criteria. The 6 dB edge of the middle discontinuity detected in the phased array C-scan is approximately 10.5”. The right edge of the PAUT indication in the C-scan is at approximately 9.25”. Therefore, these two indications are also within the 1.5” window. As a result, what started out as a single discontinuity at 10.75”, with a length of 0.75”, morphed into a larger discontinuity 3.75” in length. Further analysis, based on the length rules, has failed this discontinuity.
The depth and height of the discontinuity may be analyzed in the phased array sectorial scan (S-scan). The phased array data was generated using 45 – 70 degree focal laws at 1 degree resolution. The current focal law being analyzed in the S-scan is the 53 degree angle. The depth of the defect detected by PAUT is approximately 0.20 inches from the top surface. The height of the discontinuity is measured using the 6 dB method and the Omni-scan MX-2 ultrasonic axis measurement cursors.